Good day and thank you for standing by. Welcome to the daridorexant phase III program published in The Lancet Neurology conference call. At this time, all participants are in listen-only mode. After the speakers' presentation, there will be a question-and-answer session. For participants to ask a question during the conference, please press star and one on your telephone. If you require any assistance during this conference, you can press star followed by zero. This conference call is being recorded today, Thursday the 20th of January 2022. I would now like to hand over to the speaker for today, Mr. Andrew Weiss. Please go ahead, sir. Thank you, operator. Good morning, good afternoon, everyone. This is Andrew speaking, Investor Relations at Idorsia. We're here to have a webcast to discuss the newly published results of daridorexant in the Lancet Neurology publication that was made available this morning. With me on the call today are our CEO, Jean-Paul Clozel, our Chief Scientific Officer, Martine Clozel, our Head of Global Clinical Development, Guy Braunstein. Joining us afterwards for the Q&A, we'll also have Simon Jose. Next slide. As customary, we will be making forward-looking statements during this webcast, so everybody has, with this slide, been adequately reminded of the risks and benefits and also possibilities of us making forward-looking statements in this call. Next slide. Jean-Paul, the floor is yours. Thank you, Andrew. We are on slide three. In 2022 had a very busy start. I think it's because we are now fully functional and all the teams that we have in the U.S., in Europe and in Japan are delivering on their promise. I have to say, it's fantastic to see how people are motivated and are working very hard to get this result. Next slide. Today, it was not planned, but a few weeks before the due date, Idorsia received the Japanese PMDA approval for clazosentan, called brand name is PIVLAZ, for subarachnoid hemorrhage. Unfortunately, we do not have time to discuss this very good news. It's very important, congratulations to the Japanese team. We are going to focus today on the results of daridorexant published in The Lancet Neurology. Our intention today is really to recall how did we really discover daridorexant and why we are having these fantastic results. I think that now we can go to the next slide. Next slide. Then Martine will tell you the very long way to discover daridorexant before we go to the results. Martine, please. Thank you, Jean-Paul. Good morning, good afternoon to everyone. It's a great pleasure to be with you today to announce the publication in The Lancet Neurology of this seminal paper on the results of the two pivotal phase III studies of daridorexant just over 10 days after its approval by the FDA. Next slide, please. It was indeed a long road to daridorexant. In 1998, just one month apart, two groups published on the discovery of a new peptide, a neurotransmitter, which the group of Stanford called hypocretin and the group of Masashi Yanagisawa at that time in Dallas called orexin. It happens that just one week before this publication, I had invited Masashi Yanagisawa to our basically newborn labs at Actelion, just a few months after its startup. Masashi Yanagisawa had been the discoverer of endothelin, and we had discovered the first endothelin receptor antagonist. At that time, we immediately started a program because what Masashi was describing was, he was explaining me that one week after he was going to publish in Cell on a new peptide, and what he was telling me was very particular. We immediately started a program to try to find out what was the role of orexin and what could be a potential profile, if interesting, of blocking the orexin system. It took almost one more year for Emmanuel Mignot and Masashi Yanagisawa in two different papers to show that orexin was playing a role in wakefulness. In 2007, we published in Actelion, in Nature Medicine, that indeed dual orexin receptor antagonists induce sleep in the lab and in man. Our tool at that time was almorexant. We felt that almorexant was not yet the perfect candidate. We designed a scientific product profile, which would be a better compound to take into account all of our learnings of what was the profile of dual orexin receptor antagonist, but to improve from almorexant. It took many years of organic chemistry and interactive cycles to arrive to a compound which was fulfilling the scientific product profile, and this optimized candidate was daridorexant. Entering to man, initiation of phase III, 20 years after the discovery of orexin. Today it's a real pleasure to announce the description of this phase III pivotal program in The Lancet Neurology. Next slide, please. Indeed, orexin is very particular in that, and that was one of the main reasons for which we started the discovery program even before the publication of the paper describing hypocretin and orexin. Very, very few cells in the brain produce orexin. Only about 70,000 neurons produce orexin in man, and these neurons are located in the lateral or posterior lateral hypothalamus, ONs. That was really suggesting, because the hypothalamus is a very particular and important organ for the regulation of many functions, that there was perhaps an important role of orexin. Indeed, when we discovered our first dual orexin receptor antagonist, there are indeed two orexin receptors, OX1 and OX2, we learned that there was a very specific profile of sleep induction. This sleep was not modified as compared to a normal sleep. There was no respiratory depression. The sleep was surmountable in that one could easily wake up from that sleep and not sleep if there was no pressure to sleep. In particular, upon awakening, there was no decrease in muscular strength in contrast to GABA agonist. We realized also that chronic administration was not inducing any loss of efficacy and that there was no abuse potential. Next slide nine, please. Slide nine. Next slide. Next slide. Okay, I will continue with the slides. Almorexant was not a perfect drug, as I said, and we continued our drug discovery research. Can you go to the next slide? We continued to try to combine the properties of dual orexin receptor antagonist, but with an optimal profile for a sleep drug. Indeed, we wanted to combine this precision mechanism of action, which is very different from a global sedation of the brain as that which can be obtained with GABA receptor modulator. We wanted to combine this precise mechanism of action with an optimized pharmacokinetic profile. We tried to make a next generation dual orexin receptor antagonist, which would have an optimized pharmacokinetic profile for the right duration throughout the night without residual next morning sleep effect. Hoping it could provide an improvement in something which is very important symptom of insomnia, which is impairment of daytime functioning. We were really hoping we could improve daytime functioning. We defined with these properties what we wanted in a next generation DORA, and this scientific target product profile was indeed put in motion. Next slide, slide 10, please. Our concept was that we should really try to cover about 6-8 hours of the night with an optimal drug profile for a sleep drug in insomnia. Next slide. Indeed, many drugs are having a duration which is too long, and that is normal because it's difficult to act for the duration of the night and not have as a consequence, a residual effect the next morning of somnolence. Next slide. To avoid this residual somnolence, what can happen is to obtain a suboptimal efficacy during the night in order to avoid the next day somnolence effect. Our goal was to try to have an optimal duration of action during the night while avoiding next morning residual effect. Next slide. The result of that is daridorexant. Practically, the way we did was to try to have a drug discovery effort where we would try to predict the duration of action in man of our molecules. For predicting the human duration of action, we predicted by PBPK modeling the human pharmacokinetics of our molecules. We estimated which orexin two- receptor occupancy was needed for sleep. OX2 is the receptor which is the most important for inducing sleep when one blocks the orexin system. The combination of the prediction of human pharmacokinetics and orexin-2 r eceptor occupancy needed for sleep or awakening was allowing us to calculate what was going to be the duration of action at optimally effective doses for our different molecules during the discovery process. The goal was to continue the discovery program until we would find a compound which would have an estimated human pharmacokinetics giving sufficient receptor occupancy for six to eight hours. You can see here on this pharmacokinetic slide of phase I of daridorexant, is that after an administration of the compound, there is a maximal concentration which is reached rapidly and which declines quite rapidly, achieving 24-hour extremely low levels. At eight hours, we have lost already about 80% of the plasma concentration, 70%-80%, and at 24 hours, the residual concentration is very low. Daily administration for four days shows that the next full pharmacokinetic curve is similar, superimposable to the first one. There is practically no accumulation, which allows to have a day without really significant effective concentrations any longer. Next slide. This important aspect of the scientific product profile was not alone. There were other elements in the SPP, the desired product profile, to discover a molecule with optimized efficacy and safety. This is why it took more than 25,000 molecules after almorexant in the project to arrive to one molecule. The sequence here on this slide is given by the cascade of tests in the program, and you can see how we reduce progressively by iterative cycles. We eliminated many, many of our molecules because besides the optimal duration of action which we desired, which was modeled by use of the predicted pharmacokinetics and pharmacodynamics, hoping that there would be a good effect on night variables without next morning residual effect. We had a number of additional requirements for designing the optimal molecule. A high inhibitory potency on both orexin -1 and OX2 receptors. Indeed, in addition to antagonism of OX2 receptor, fundamental for sleep, we felt important to find the molecule which would block as much the orexin- 1 receptor. Indeed, OX1 mediates the link between orexin-producing neurons and systems involved in anxiety, mood, hyperactivity. Therefore, we felt that OX1 blocking was important as well to help solve the fear of not being able to sleep, characteristic of insomnia, and hopefully help to solve also the daytime impairment in addition to the improvement in night variables. We, of course, wanted also a high and fast brain penetration to help efficacy and speed of action and also various aspects to optimize safety. Next slide, please. 15. The result of this more than 20 years of drug design is a compound, I think, with outstanding qualities. Indeed, the preclinical characteristics of daridorexant were combining an equipotent antagonism of both orexin receptor, high brain penetration, a rapid absorption and a moderate duration of action of about eight hours in insomnia, a very safe compound in toxicology with no teratogenic or carcinogenic potential, a chronic efficacy, and also a preservation of muscular strength and a preservation of memory in the lab upon awakening with no abuse potential. I would like now, next slide, 16, to hand over to Guy for the presentation of the data published today in The Lancet Neurology of the two pivotal phase III studies of daridorexant. Guy? Thank you, Martine. Good morning. Good afternoon, everybody. Slide 17. We can move on immediately even to slide 18. On slide 18, you see immediately the conclusion from the Lancet Neurology article that daridorexant 25 mg and 50 mg improved sleep outcomes, and daridorexant 50 mg also improved daytime functioning in people with insomnia disorder with a favorable safety profile. What I'm going to do in the next 20- 30 minutes is to tell you how we got there. Of course, as you can imagine, there is a wealth of data and results in the publication and in the supplement, and I invite you to read in detail the paper and the supplement because I won't be able to show you everything that is already published. Next slide, please. Before I jump into the study itself, I would like to maybe remind you a little bit about the disease insomnia, and first to mention that insomnia is primarily a subjective patient experience. I'm saying that because we place patients in the center of our development program. Aligned with the DSM-5 definition, we know that insomnia affect both night and day. This is also important to us because, as already mentioned, we want to show the efficacy of the product on the night symptoms as well as the day symptoms. From patient input and patient research that we have conducted, we have identified daytime symptoms and total sleep time as the major concern to them. This is important also in terms of the choice of the endpoint that I will describe later on. Finally, I would like to mention that in most of the development programs of hypnotics, daytime symptoms have been largely ignored. Which means that in the end, we know very little about impact of currently available products on the daytime symptoms of patients. This may be due to the desire to ignore it, or maybe most likely as well, the fact that there was no tool, no instrument, patient-reported outcome instrument to collect properly the impact of products on the daytime functioning of patients. Next slide, please. Here is a schematic of the design of the development program, phase I, phase II, and phase III. What I would like to mention here is that at the end of phase II, we knew quite a lot about daridorexant. We knew that there was a dose response efficacy on nine time parameters, WASO, LSO, LPS, total sleep time. We also knew, and that was very important, that the full night was covered with the effect of WASO increasing quarter by quarter and culminating the last quarter of the night. What was also very interesting is that from the phase II study, and even from phase I, we knew that there was no persistent effect beyond eight hours, and when patients wake up in the morning after a night of dosing, there was no sleepiness in the morning. Finally, we knew at that time that there was a signal on daytime functioning that was with a crude instrument, a visual analog scale only, but we had the idea that the profile that Martine described before could actually be reached. However, there were a number of open questions still. The first one being the dose that can provide the best balance between a good night, a good day, without morning sleepiness. It's clear that there is a fine balance here between the good effect and the potential hangover effect, as shown by Martine on the pharmacokinetic profile of the product. It's not easy to determine the dose that will provide the best benefit risk from that perspective. This explains why we had to test several doses in the phase III program, and we decided to test the 10, 25, and 50 mg. The other aspect that was still open was to have a proper measure of daytime functioning. We knew that there were no adequate tools that will allow us to properly measure the daytime functioning from a patient perspective, and we had the chance to develop this instrument in parallel to the development program of daridorexant, including the phase II study. Finally, we could design the phase III program through the doses, as I mentioned, 10, 25, and 50 mg. We looked at the effect during the night and the morning, the effect during the day. Of course, the safety with long-term treatment was absolutely essential, as well as the maintenance effect with continuing treatment. All of that is actually addressed in the phase III program. Next slide, please. We are on slide 21. The key question here is, how can we measure daytime functioning? As I mentioned, there was no tool available, and we embarked into the development and validation of a new instrument called IDSIQ. The basis of this instrument was in the publication in 2007, called the DISS, Daytime Insomnia Symptom Scale. This was a good baseline to develop a better instrument, but it was insufficient from a patient input perspective. During the development of the final instrument, IDSIQ, we had a series of patient interviews that allowed to elicit concept to ensure saturation. We had also cognitive debrief that allowed to make sure that the patients would understand the questions, be able to respond to them, and eventually validate the content of this questionnaire. Following this development, the validation was done through a series of studies, two studies, one in healthy participants and another one in insomnia patients. These studies were conducted in parallel to the phase II program of daridorexant. The statistical analysis validated psychometrically the instrument. In particular, we could show that there was a very different response between healthy subjects and insomnia patients, as well as all the characteristics culminating in the final IDSIQ questionnaire, which was eventually accepted by the FDA as appropriately validated instrument. We had several interactions with the FDA and confirmation that that was an instrument that would fit the FDA needs. One important element of the validation was to identify the within-patient meaningful difference, and this is actually published in a peer-reviewed journal in 2021. All the details of this development are actually published. Maybe one point to mention as well is that this questionnaire was translated into multiple languages, which was useful for inclusion into our program that was multinational. Now that you know how we did it and that this really took several years to maybe six, seven, eight years of work to develop this instrument, we can look at the instrument itself on the next slide 22. So here you see on the left side of the slide, the exact questions in the right order that were raised to the patient every day. We see the variety of the diversity of questions here, and as you see, the phrasing is really coming from the patient themselves. We see the diversity of the questions can go from being able to concentrate, being frustrated, irritable, maybe how much of an effort was it to perform daily activities, as well as being tired, mentally or physically tired, sleepy. You can see all these questions. These questions are addressing three key dimensions of daytime functioning. The first one is Cognition, and the domain score is called alert cognition in blue, made of six question items 1, 2, 3, 9, 10, and 14. The second one, which is called the Mood domain, is really looking at the affective dimension of daytime functioning, and it's made of four questions. Finally, the physical component of daytime functioning is assessed by the Sleepiness domain score, which is made of four questions. Item 8, how energetic did you feel today? Item 11, how mentally tired did you feel today? Item 12, how physically tired did you feel today? And item 13, how sleepy did you feel today? On the right side of the slide, you see a typical screenshot of what the patients were actually seeing when they were entering every day the response to the questions. It's important to mention that this is a daily assessment. The recall period was today, and this avoids the problem of memory bias when the recall period is longer. Now that you are very familiar with the instrument that we have, I can go to the next slide 23, which shows the design of the study, very classical. A screening phase with a double-blind placebo run-in phase of two to three weeks, during which the patient had a polysomnography. The patients were randomized to either a high dose, low dose, or placebo for three months with an interim look at the first month visits, with two polysomnography nights. At the end, at three months, again two polysomnography nights. Finally, a single blind placebo run-out period designed to assess a rebound effect, if any, the withdrawal symptoms. Eventually, the patient could move on to the extension study, so that in total they could be treated for one year, or maybe they didn't want that, there was a safety follow-up period. We did two identical studies, study 1 and Study 2. Study 1 had the high dose of 50 mg and the low dose of 25 mg, Study 2 had the high dose of 25 mg and the low dose at 10 mg. By doing that, we had two arms that were similar in both studies, the 25 mg and the placebo, and two arms that were different in the two studies, the 10 mg Study 2 and the 50 mg Study 1. next slide, please. Slide 24. Here we see the key assessments. As I mentioned before, we assess the polysomnography at baseline at one month and three months of treatment, as well as during the single-blind placebo run-out period. This was needed for to meet regulatory requirements. There is really an appetite from the regulators for objective sleep measures. We also wanted to have patient-oriented outcome every day, because as we know, and as I remind you earlier, insomnia is primarily a patient subjective experience. We had two categories of patient-oriented outcome instrument or measures. One was the night with the Sleep Diary Questionnaire that I will come back to in a moment, and the IDSIQ questionnaire, which I described before. The next slide shows you the Sleep Diary Questionnaire in brief. Again, that was a daily recording. This questionnaire has actually two components, a morning questionnaire made of 10 questions. From those, the total sleep time is the one that was the most important. As I mentioned, this is one of the main concerns from patients. There were also three visual analog scales that were assessing in the morning the quality of the sleep, the deepness of the sleep, and how sleepy patients were in the morning. In the evening, there were additional questions for the day relating to napping during the day, as well as two visual analog scales, again related to the alertness of the patient during the day and the ability to perform their normal tasks during the day. Next slide, please. Slide 26 is now showing the key endpoints that we had in the study. As I mentioned, we have objective and subjective endpoints. The two primary endpoints were derived from the polysomnography, the Wake After Sleep Onset, or WASO, and the Latency to Persistent Sleep, or LPS. We had two secondary endpoints derived from patient questionnaires, the night and the day. The first one was subjective total sleep time by the SDQ that I showed you before, and the sleepiness domain during the day by the IDSIQ questionnaire. As mentioned, we had two main endpoints, two main time points, the Month 1 and the Month 3, where all these endpoints were assessed. Of course, we had two dose levels in each study. Study 1, 50 mg and 25 mg as placebo, and Study 2, 25 mg and 10 mg as placebo. In total, when you look at all these, endpoints, time points, and doses, you end up with a need to execute 16 comparisons to placebo. Excuse me. What we needed to do is to control the Type I error at a study level, so that a solid conclusion can be made from a statistical perspective. When I presented that some years ago now, at the beginning of our program, everybody was questioning, "How can you do that? How can you actually control the Type I error across that many hypothesis testing? This is shown on the next slide. I can spend maybe a minute or two describing it because this is really crucially important. The way to do it first is to organize this hypothesis testing in a logical way. To do that, we sort of prioritize initially the high dose versus placebo, and this is represented by the green and the blue bubbles. Then the low dose represented by the red and the purple bubbles. Within each dose level, we prioritize Month 1 followed by Month 3. This is the green versus the blue and the red versus the purple. There is a reason to do that. There is no reason to look at the low dose if high dose doesn't work, so therefore we start with the high dose. There is no reason to look at three months if the product doesn't work at one month. This is why we had these four consecutive buckets here, high doses placebo at Month 1, Month 3, and then low dose of placebo at Month 1, and low dose of placebo at Month 3. However, we still have to deal with the four endpoints for each of these four comparisons. To do that, we prioritize the objective measures, whether on LPS first, followed by the subjective measures of sTST and IDSIQ. Of course, we want to control the Type I error for all of those. We start with an alpha level of 0.05, and we initially apply the Bonferroni correction. In doing that, we divide alpha by two to test the hypothesis, the null hypothesis for LPS and the null hypothesis for WASO. Depending on the results of this first comparison, which is tested at 0.025 for LPS and 0.025 for WASO, the propagation of the alpha depends on whether we can reject the null hypothesis or not. If we can reject the hypothesis, then the alpha is propagated below following the red arrows. If it's not, if we cannot reject the null hypothesis, then the alpha stop there and is in a way lost. If I take the first example of the Study number 1, when we test the first two hypotheses, H1, H2, the WASO and LPS, we could reject both of them. Therefore, the 0.025 for H1 and H2 could be preserved and propagated. The propagation was half of it, so 0.0125 was going to H3 from LPS, and similarly, 0.0125 was coming from H1 to H3, which means that in total, the sTST H3 could be tested at 0.025. Half of the alpha from H2 was going to H6, and half of the alpha from H1 was going to H5, and so on. Here you see the progression from one variable to the other. We were very fortunate in the three or one study that there was no failure of rejecting any hypothesis up to H 12. Therefore, up to that level, the alpha will stay intact. It's only H 12 that failed. I will show you later. However, H 14 and 13 could be tested, H 15 as well, and the final one, H 16 failed as well. We were very, very fortunate here that with this structure, we could actually make very robust conclusions on the high dose at Month 1 and Month 3, the low dose at Month 1 and Month 3 on most of these variables. In the end, 14 of these 16-hypothesis testing we could reject the null hypothesis and make a conclusion. This is how we did it. It's fully described in the supplement of the Lancet Neurology article, and I invite you to read the text that goes with this design and also to look at the page 19, which shows the alpha for each of these hypotheses and the P values, and you can compare the two and see whether we reject or not the null hypothesis. In addition to these 16 endpoints and measures, we had a number of exploratory endpoints, and they are described on the slide 28. Next slide, please. You see that we had also TST by polysomnography, as well as a number of variables from PSG, in particular, the percentage of sleep time by each sleep phase. We had, as I mentioned before, the morning VAS for depth and quality of sleep. We had the change in Insomnia Severity Index, a very classical instrument, ISI. We had, of course, the IDSIQ other domains, the mood domain, as well as the cognition domain. We had the evening VAS that I described and of course, the adverse event, the withdrawal and the rebound. With all these endpoints, we have really a rich body of evidence supporting our product. Now that you understand the methodology for the development, we can start to move on to the patient demographics on slide 29. This is very classical. You have all the data in the paper. I will just point out that we had approximately 70% female, which is normal for insomnia studies. We had approximately 40% of patients, elderly patients, which is very important in that indication. Interestingly, on the bottom, you see that we have quite a large number of obese or overweight patients, and that's of course normal for insomnia population. In total, we had really a population reflecting well the real-life situation. The baseline characteristics are on the next slide, and we see that slide 30. What we see here is that the WASO approximately was an hour plus 35-40 minutes. The LPS was approximately one hour, and the total sleep time was very short, around five hours. This really reflect a well-defined insomnia population. Next slide, please. From the set of results that are published in the Lancet, I cannot, of course, in the next minutes, show you all the data. Because the results of the 10 mg didn't provide benefit, and the 25 mg was very consistent between the two studies, I decided for the purpose of this presentation to focus Study 1. the rest you can really see, everything is presented in the publication, the supplemental. We now just concentrate Study 1, the placebo 25 and 50 mg doses. Next slide 32. Let's stay with me for a second that I can describe how the slide is constructed, because we'll use the same format for several other endpoints. On the right of the slide, you see the graph with the observed values. The baseline, the Month 1 and the Month 3, with three lines, the placebo in purple, the 25 in blue, light blue, and on green, the 50 mg. On the bottom right, what you see are the statistics, the effect of placebo, the effect of daridorexant 25, of daridorexant 50, as well as the placebo-corrected differences. On that one, you see the change from baseline. On the graph, you see the values, and on the table, you see the change from baseline and the placebo-corrected change from baseline. Then you have this red box, which basically provides the conclusion. As we can see immediately here, we have a nice effect on WASO from daridorexant. The P values in the rest of the presentation are only shown for those that are statistically significant. We see a highly statistically significant effect of 25 and 50 versus placebo. This is what we concluded at Month 1 and Month 3. What we see as well is that the improvement is achieved at Month 1 and stay the same at Month 3, with no further improvement. On the next slide, using slide 33, using an identical format, we see the results for LPS. We see the very nice reduction. The placebo effect seems to be more important here than on WASO, but we see an improvement by approximately 30, 35 minutes on daridorexant 50 mg. We see the high level of significance with the P values. We again see that the effect is at Month 1 and remains stable at Month 3. These are really confirmation that daridorexant at 50 and 25 mg are inducing sleep and maintaining sleep. Now, what about the patient feeling? This is shown on the next slide 34, where we have the subjective total sleep time. Here we see a pattern that is maybe slightly different. At, of course, now total sleep time has to increase, not decrease on the previous slides. We see an increase of the total sleep time. You have the numbers here on the table, and you have the absolute value in the graph. What we see here is that the effect at Month 1 is good, but there is even more effect later on at Month 3. Eventually on 50 mg at month three, we see that the increase in total sleep time is reaching an average approximately one hour, which is considerable for these patients. Remember that at baseline, they sleep for approximately five hours, and now on the treatment, they sleep at least six hours, which is really an achievement. As we see, the conclusion is that daridorexant 25 and 50 mg are significantly impacting subjective total sleep time compared to placebo at month one and three. Not only the patients are sleeping better in the sleep lab with the PSG, but also feeling this improvement in total sleep time. They are able to report it by themselves. There are a number of other analyses done during the night. I can only summarize them on slide 35. As you see here, the proportion of sleep time spent in each sleep stage is preserved. daridorexant is increasing the sleep time for the REM and non-REM sleep identically. There is no preference for one of these stages, so the sleep architecture is preserved. We see in the middle that the quality of sleep and the depth of sleep are actually numerically higher on daridorexant versus placebo at month one and month three, which is really good. Again, it's a perception by the patients that the night was actually good, and the ISI is also moving in the right direction. I want to point out that there was no statistical testing on these endpoints, but numerically, they are all in favor of daridorexant. Let's move now on the daytime functioning, and this is shown on the next slide, which has an identical format, slide 36, an identical format to the previous slide. We see the graph on the left with the actual measures, and you see the statistics on the right and the conclusion. Here, what is really interesting, again, is that from a patient perception during the day, there is an improvement at Month 1, but it continues to improve over time at Month 3. The pattern of patient feeling, as we have seen for sTST and now for IDSIQ, is actually slightly different from the PSG. Here we could make the conclusion that 50 mg significantly improve IDSIQ sleepiness domain. However, 25 mg was not statistically significant, and that's why the P value is not provided. However, if you look at the paper, you will see the P values for all these endpoints as well. Looking now on slide 37, I'm here reporting the results for the other domain. Again, no P values because although you can see them in the paper, they were not adjusted for multiplicity. So, for simplification, I don't provide them, but you can see them in the supplement. We see identical pattern of improvement with the placebo effect being relatively modest, with the 25 mg getting better and the 50 mg clearly differentiated from the 25 and the placebo on the Mood domain, the Alert/Cognition domain, as well as the total IDSIQ score. Here I want just to point out that there is consistency, at least numerically, on these different endpoints derived from the IDSIQ questionnaire. Finally, just to bring you a few more results on slide 38. Again, during the day, we had this VAS, visual analog scale, scores for ability to function on daytime alertness. This again, numerically were better at Months one and three on daridorexant versus placebo. We had this morning sleepiness VAS score, which indicate less sleepiness in the morning in the patients with the daridorexant compared to placebo at Month 1 and Month 3. Again, these are not statistically tested because they are not in the hierarchy, but the trend and the numerical values are very clear, and you will be able to see all these results again in the supplement. Finally, to finish on the efficacy results, I want just to point out on slide 39, that the effect on WASO, LPS, as well as total sleep time on sTST were very consistent between the different subgroups, adult versus elderly subjects across sex, male and female, as well as geographical location. You see here just one example of WASO at Month 3. But all the data for all the endpoints at Month 1 and Month 3 are actually presented in the supplement. You will see by yourself that they are very, very consistent between subgroups. Of course, I cannot stop here. I have to talk a little bit about the safety of the product. I will start by and maybe remind first that for transparency, I will show you the result, the safety results Study 2, although I focused the efficacy Study 1. next slide 40, shows the adverse event profile. We see that there are actually relatively few adverse events in general. Only a third of the patients have reported one or more adverse events, and this is very balanced between the treatment groups. We see that the adverse events leading to treatment discontinuation is very rare. Actually very few of them. We can more or less count them by hand. And we see that the most frequent adverse events are really non-specific, nasopharyngitis or headache. Maybe it's more interesting to look at the somnolence with 2% on placebo in Study 1 and 1% Study 2. You see a number that are really very close to that in Study 1 and Study 2 on daridorexant. 2% on 50 mg, 4% on 25 mg Study 1, and 3% on daridorexant 25 mg and 2% on daridorexant 10 mg in Study 2. We have really a very small difference between daridorexant and placebo. Maybe it's time to also point out that from a safety perspective, the profile of 50 doesn't really differentiate from lower doses. We can see also the values for fall or fatigue or dizziness, and all of that is extremely consistent with very small number of adverse events reported from daridorexant. The next slide 41, is a focus on adverse events which might be seen as maybe more concerning, excessive daytime sleepiness, sleep paralysis, hallucination, suicidal ideation, or self-injury. We had very few of them. We had an adjudication committee independently, blindly reviewing these adverse events. You see the data here, they are really very, very rare. They're important, but they are rare and somewhat balanced as well between daridorexant and placebo. Finally, on the adverse event description, I would like to point out the effect of gender, sex, sorry, the effect of age, the younger and the older patient, and you see that on slide 42. On the left-hand side of the slide, you see the younger patients. On the right-hand side, you see the older patients. As you can see, it's very, very consistent and even sometimes, maybe less adverse events in elderly compared to the younger. Of course, these are small numbers, so it's difficult to make a real conclusion. If you look at somnolence, you have 2.1% on 50 mg in the younger, and you have only 4.8%, actually one patient only in the elderly. If you look at the fall, which is one of the main concern of hypnotic in the elderly patients, you see actually on the right side that there were more, numerically more patients having falls on placebo than on daridorexant. These are really reassuring findings that the safety and tolerability of daridorexant even at 50 mg and in the more fragile elderly patients is very good. We had further safety observation that I would like briefly to mention, but all the details again are in the publication. Slide 43, we see no adverse event suggesting the drug misuse, which is an important observation. We see that there was no suggestion of withdrawal symptoms during the run-out phase. There was also no rebound detected during this placebo run-out phase. Just to finish with a couple of slide on the description of the paper, there of course is slide 44, some strengths and limitations of the studies. On the left you see the strengths with the assessment of most components of insomnia, in particular, objective and subjective endpoints, the night and the day, which is unique to this development program. We had robust Type I error control, which is also very important to make robust conclusions. We see that the description of patient is really reflecting the real situation with more female than male and a nice representation of elderly patients, approximately 40%. Something I didn't mention before, but we had a very low rate of missing data and actually also very low rate of discontinuation of patients. Of course, as normal, we have to balance that with limitations. There are a few that are mentioned in the paper, which we can talk briefly about. Mainly, the study focused on moderate to severe insomnia. This is reflected by the ISI score above 15, as part of the screening values. As we know, below 15 it's borderline insomnia. This reflects really the real-life situation. Regarding race, we may have had maybe underrepresentation of minorities. Maybe more interesting is to see that we had few patients with CBT-I experience. Again, this is reflecting the real-life situation that although CBT-I is recommended as first line, there are actually very few patients that have the time, the energy, or the possibility to find the psychotherapist. There is also a price constraint, so it's difficult to do CBT-I and therefore the fact that we had only few patients with CBT-I reflects the real-life situation. Of course, comorbidities, we didn't have many, but there are reasons for that that are fully described in the protocol, which is in the supplement. Finally, the issue of having a new instrument like IDSIQ makes, of course, the discovery interesting but more surprising and difficult to interpret for some people. In the end, the next slide is showing again the conclusion from the paper and the fact that there are requirements for insomnia treatment these days, that is to improve both nighttime and daytime symptoms, and that we really need to prevent the risk of next morning residual effect. It seems that, and that's what is written in the paper, that daridorexant 50 mg seems to fulfill these requirements. We have improvement on sleep variables without excessive sleepiness and good effect on daytime functioning. Next slide, please. I would like now to maybe give more a kind of Idorsia conclusion on this development program, and this is shown on the next slide 47, which will be my last slide. We have a comprehensive sleep efficacy and perceived by the patient. Patients fall asleep faster, they still sleep longer, and the total sleep time was increased, perceived by them, and that was with preservation of the sleep architecture. We also have daytime functioning perceived by the patient with formal statistics on the sleepiness domain, as well as consistent efficacy on the other domain, although not statistically tested. Interesting is to see that this effect is maybe progressing over time, and maybe not observed to the maximum effect very shortly. Finally, the safety is good. The low, very low, incidence of adverse events comparable between groups. Very few serious adverse events. I didn't mention them, but very, very few. There was no evidence of tolerance or dependence and no rebound. All of that is probably due to the mode of action as described by Martine before, the precision mode of action, as well as the intrinsic properties of daridorexant. It's a targeted mechanism and an ideal pharmacological profile, which allow us to have all these conclusions, and they are not automatically true for other products, but certainly for daridorexant. With that said, I will stop. I see that it's quite late now, and maybe take a few questions, if any. Okay. Thank you, Guy, for your concluding remarks. We have come to the end of our prepared remarks and these slides and are ready to take the Q&A session full frontal. As this is a scientific call, please, I want to remind everybody that please address your questions to this topic. We will have a financial call for the full year results in about three weeks' time and would defer those kinds of questions to then. Operator, please open the lines. Thank you. As a reminder, if you wish to ask a question at this time, please press star one on your telephone and wait for your name to be taken by an operator. Once again, it's star followed by one if you have any questions or comment at this time, and it's the pound or the hash key to cancel. We have our first question coming from Peter Verdult at Citi. Please ask your question. Thank you. Pete at Citi. Thanks to all of you for hosting the call. Now given how thorough your presentation was on the science and the clinical data, it would be difficult for me not to ask maybe a commercial question or two. Can I just kick off with Guy? Apart from supporting, you know, the Alliance for Sleep and investigator-led studies, are there any plans for, from an Idorsia point of view to conduct some sort of phase IV program as it relates to Dari? Just looking into your early-stage discovery efforts, are there any ongoing research efforts in insomnia or new mechanisms or targets that you would flag at this juncture? I'll pause there and then I have just two very quick follow-ups. Thank you. Okay. Thank you, Peter. Guy, you want to take this phase IV? Yes. I can maybe start the answer to the question. Of course. With this, there are two things I can mention. First, we are not at the end of the full publication of the data we already have from the trial. You will see a lot more data coming gradually as we publish. We also want to do more studies. With a product of that quality, there are many areas of research that we are considering and working very diligently on creating new studies so that we understand better the efficacy and the safety of daridorexant within the indication, maybe in some subgroups. This is work in progress, and I'm sure you will hear about that in the near future. Okay. Thank you, Guy. Martine, do you want to comment on mode of action or other insomnia products that we may have in early development? Yes. We have a phase II running at the moment in an indication of binge eating with a selective orexin- 1 receptor antagonist, which should have a profile of a compound without sleep effect. Binge eating is a big medical need. We will have the result this year. Beyond, it's too early to speak, but of course, we are interested in continuing in the fields on which we have been working. So far, I think, it's not yet visible. It's a lot of research which has gone into orexin and insomnia understanding. Thank you. Thank you, Martine. Thank you, Andrew. Forgive me, if I may just squeeze one last one in just for the team. Look, I know you said you're not going to talk about financials, but I would like. Could you now you've got the label in hand. You've got Jennifer Aniston on board. You know what you're dealing with. You've got The Lancet Neurology paper. I mean, the market's penciling in $1 billion, a blockbuster status. But when you as a team talk about daridorexant, could you at least give us a hint as to you know, what your expectations are and what you think, how big you think this drug can be? Thank you. Thank you, Peter. Jean-Paul, do you want to take more of a strategic view on this? We can give maybe Simon the opportunity to comment on more details on marketing and tactics here. Jean-Paul, you may be on mute. I just wanted to say simply, sky is the limit. You know, frankly, we have an extraordinary drug for millions of people. I think that, as I said, you know, this drug will ensure growth for the company for the next 14 years because it's a very long patent life. We will have to work very hard to gather more and more patients. We are going to give, in a few weeks, our first vision for the beginning of the launch. Really, since I always said that zolpidem Ambien used to sell for $4 billion 15 years ago, we should do better, but it will take time. Thank you, Jean-Paul. Simon, do you want to try to stitch together- Sure. ... How you see the nearer term maybe, and how- Yeah. ... are we putting tactics together with all of these elements that we now have laid out since the approval about 10 days ago? Yeah, I mean, I just maybe to add a comment to Jean-Paul's. I mean, just to give you a sort of numerical framework. If a lot of people are talking about price, and if you start with suvorexant price as a sort of benchmark. At a suvorexant price, 3% market share in the U.S. alone is $1 billion at list price. So, you know, you can look at the label, look at what we've got and work out what sort of penetration rates we can get. But for every 3%, you get about $1 billion in the U.S. at list price. So as Jean-Paul says, you know, with the size of the market and the profile of the product that we have, it's very significant. Then on top of that, we are actually very excited about Europe. We'll be the first dual orexin receptor antagonist in Europe. There's a lot of excitement. There's been no innovation in the European insomnia market for 20-odd years, and we're finding a great deal of receptivity toward the profile as we get into our discussions about Europe. I think that's something else that I think you can add on to the opportunity in the U.S. Very clear. Thank you. Thank you, Simon. Thank you, Peter. Operator, next question, please. Once again, ladies and gentlemen, if you do have any questions or comments at this time, please press star followed by one on your telephone. We have the next questions coming from the line of Manos Mastorakis from Deutsche Bank. Please ask your question. Yes. Hello. Thank you for the call, and thank you for taking my question. Essentially, I just wanted to ask the extent to which you will be using this Lancet publication in your marketing efforts and discussions with physicians, all these engagements? Thank you. Thank you, Manos. Simon, do you want to provide any context- Yes, sure. Yes. sure. Yes. ... as to how these fits in our ability to market? Yes, sure. Thank you, Manos. It's critical, frankly, because obviously the depth of data that we're now putting in or putting out there through the publication is much broader than we have in the label, which is often the case. But when you see all the IDSIQ data, now in this publication, including the exploratory endpoints, the VAS, and I think really critically, the somnolence data. T o be looking at, there's no difference between doses and somnolence, not that much difference to placebo. And as Guy said, strikingly, you look at the elderly population, you've only got 0.8% somnolence. All of these data are going to be important. They are now in the public domain, and we do expect to be using them proactively and disseminating these data through our different communication channels. Thank you, Simon. Operator, next question. We have no further questions at this time, so please continue. Okay. Operator, thank you very much. Jean-Paul, do you want to make some concluding remarks? No, I want to thank Guy and Martine, and I just like to repeat that it took us really 25 years to get there, but it was really worthwhile. I think the data are really frankly. If at the beginning of the program, we would have said we would have this type of data, we would not have believed it. I think that it shows that to have looked for the ideal profile really made a difference. I repeat, if you look at the difference within the same drug between 25 and 50, you understand that this program really, and the drug, have been really allowed us to to really discover something which is going more than to improve many patients. I think we are dealing with insomnia with a society problem. There are more and more insomnia. We have seen that with COVID, there are 20% more patients than usual who are claiming that they cannot sleep or who are really a problem. I think that with this tool, with this drug, we are going to be able to change the life of million and improve the life of millions of patients. This is very exciting, and I hope that I invite everybody to look at this data, to go into the details. It's a lot of data, but it's worthwhile to really make your opinion yourself. I hope that we have helped everybody to understand better this paper and this development program. Thank you, Jean-Paul. We have passed the top of the hour, so it is time to conclude our webcast. Thank you for your ongoing interest in Idorsia. It is an exciting year that we're in right now. Our next scheduled press release is for the full year 2021 results, scheduled for the 8th of February. Until then, have a good time. Operator, please close down the lines. Thank you, ladies and gentlemen. That does conclude our conference today. Thank you for participating. You may now disconnect your line. Speakers, please stand by.